scientific correspondence
that all the crossbridges in a rigor muscle
contribute to the stiffness. If this is not so
— for example if many are bound in ‘slack’
states — the calculated duty ratio would be
lower and the discrepancy might vanish. If
the duty ratio is indeed small, the experi-
ments of Higuchi and Goldman8 are consis-
tent with a small working distance. The
experiment of Yanagida et al.11 remains dif-
ficult to explain.
Joe Howard
Department of Physiology and Biophysics,
University of Washington, Seattle,
Washington 98195, USA
e-mail: johoward@uwashington.edu
1. Howard, J. Nature 389, 561–567 (1997).
2. Finer, J. T., Simmons, R. M. & Spudich, J. A. Nature 368,
113–119 (1994).
3. Ford, L. E., Huxley, A. F. & Simmons, R. M. J. Physiol. 269,
441–515 (1977).
4. Huxley, A. F. & Tideswell, S. J. Musc. Res. Cell Motil. 17,
507–511 (1996).
5. Eisenberg, E., Hill, T. L. & Chen, Y. Biophys. J. 29, 195–227
(1980).
6. Ford, L. E., Huxley, A. F. & Simmons, R. M. J. Physiol. 361,
131–150 (1985).
7. Yanagida, T., Arata, T. & Oosawa, F. Nature 316, 366–369
(1985).
8. Higuchi, H. & Goldman, Y. E. Nature 352, 352–354 (1991).
9. Huxley, A. F. Prog. Biophys. Biophys. Chem. 7, 255–318 (1957).
10. Lovell, S. J., Knight, P. J. & Harrington, W. F. Nature 293,
664–666 (1981).
11. Yanagida, T., Arata, T. & Oosawa, F. Nature 352, 352–354
(1985).
A new classification
for HIV-1
The phenotype of HIV-1 isolates is defined
by the cells in which they replicate in vitro,
but these phenotypes can change in vivo
with profound implications for viral trans-
mission, pathogenesis and disease progres-
sion. Here we propose a new classification
system based on co-receptor use, providing
a more accurate description of viral pheno-
type than the present imprecise and often
misleading classification schemes.
At present, three classification systems
are in use. The first defines primary isolates
as macrophage (M)-tropic or T-cell-line
(T)-tropic. However, this system disguises
the fact that all primary isolates replicate in
activated, primary CD4+ T-lymphocytes.
The second system categorizes isolates as
being either syncytium-inducing or non-
syncytium-inducing (NSI) on the basis of
whether they form syncytia in MT-2 cells,
which express CXCR4 but not CCR5 (ref.
1). However, NSI viruses can readily form
syncytia with CCR5-positive cells. The third
system defines viruses as either slow/low
(SL) or rapid/high (RH) depending on their
growth kinetics in culture2. These classifica-
tions are often used interchangeably, but
they are not synonymous.
The identification of some chemokine
240
receptors as having critical roles in the cel-
lular entry of HIV-1 allows us to develop a
more precise system for identifying the phe-
notypic properties of virus strains. A major
determinant of HIV-1 tropism (phenotype)
lies at the level of virus entry into target
cells, which in turn is governed by the
expression of co-receptors in conjunction
with CD4 (refs 3–5): either CCR5 or
CXCR4, or both. CXCR4 use is a defining
feature of viruses that form syncytia in T-
cell lines; use of CCR5 is a property of NSI,
M-tropic viruses; and many T-tropic pri-
mary isolates can use both co-receptors3–5.
A nomenclature based on the co-receptor
used would thus provide a precise molecu-
lar designation of a given isolate that largely
explains its phenotype.
We propose that isolates that use CCR5
but not CXCR4 be termed R5 viruses, that
isolates using CXCR4 but not CCR5 be des-
ignated X4 viruses, and isolates able to use
both co-receptors with comparable efficien-
cy be called R5X4. Whether an X4 or R5X4
virus is a cell-line-adapted isolate should
also be specified.
Under this system, R5 viruses are the
strains most commonly transmitted sexual-
ly, consistent with the high resistance of
individuals lacking CCR5 to infection6,7.
After about five years, viruses evolve in
about 50% of patients that are able to use
CXCR4, with or without concurrent use of
CCR5 (refs 8,9). These viruses would now
be called R5X4 and X4 viruses, respectively.
Isolates passaged through a permanent T-
cell line should be called T-cell line-adapted
(TCLA) X4 or R5X4 viruses, and a similar
qualifier can be used for viruses adapted to
growth on other cells.
This nomenclature takes note of the
ability of an isolate to use the major co-
receptors, but does not specify whether the
isolate can replicate in a particular target
cell. The nuances of co-receptor usage in
specific contexts are beyond a simple classi-
fication system and should be specified by
authors if there are perceived ambiguities.
That a virus can use a particular co-receptor
in transfected cells does not mean that this
virus uses the same co-receptor in a more
physiological context. The efficiency with
which different co-receptors are used by
some strains is likely to vary between assay
systems; again, authors should clearly
explain the limitations of their results, and
the significance of extremely low efficiency
co-receptor usage should not be over-inter-
preted. This classification system can also
be expanded to take note of other co-recep-
tors if their use by an isolate proves to be a
major determinant of tropism; for example,
whether isolates use CCR3 and/or CCR5 to
enter microglia could define them as R3, R5
or R3-R5 isolates10.
Our classification system is uncompli-
cated, is intuitive to those familiar with the
Nature © Macmillan Publishers Ltd 1998
usage of the CCR5 and CXCR4 co-recep-
tors by HIV-1, and removes the inaccura-
cies and confusion associated with the
present systems. It is flexible and open to
expansion to accommodate emerging
knowledge of co-receptor usage, and can
encompass HIV-2 and SIV strains, which
also use CCR5 and other co-receptors for
entry3–5.
A record of co-receptor use by particular
HIV-1, HIV-2 and SIV strains will be main-
8
tained by the Los Alamos National Labora-
tory Sequence Database, together with an
expanded version of this article.
E. A. Berger
Laboratory of Viral Diseases, NIAID,
National Institutes of Health, Building 4,
Room 236, Bethesda, Maryland 20892, USA
R. W. Doms
Department of Pathology and Laboratory Medicine,
University of Pennsylvania, Philadelphia,
Pennsylvania 19104, USA
E.-M. Fenyö
EC Concerted Action on HIV Variation,
Microbiology and Tumorbiology Center,
Karolinska Institute, 171 77, Stockholm, Sweden
B. T. M. Korber
Theoretical Biology and Biophysics, Los Alamos
National Laboratory, Los Alamos,
New Mexico 87545, USA
D. R. Littman
Howard Hughes Medical Institute,
New York University Medical Center, New York,
New York 10016, USA
J. P. Moore
The Aaron Diamond AIDS Research Center,
The Rockefeller University, 455 First Avenue,
New York, New York 10016, USA
Q. J. Sattentau
Centre d’Immunologie de Marseille-Luminy,
Case 906, 13288 Marseille, France
H. Schuitemaker
Department of Clinical Viro Immunology,
Central Laboratory of the Netherlands Red Cross
Blood Transfusion Service and Laboratory for
Experimental and Clinical Immunology,
University of Amsterdam, Amsterdam,
The Netherlands
J. Sodroski
Department of Cancer Immunology/AIDS,
Dana-Farber Cancer Institute, 44 Binney Street,
Boston, Massachusetts 02115, USA
R. A. Weiss
Chester Beatty Laboratories,
The Institute of Cancer Research, 237 Fulham
Road,
London SW3 6JB, UK
1. Schuitemaker, H. et al. J. Virol. 66, 1354–1360 (1992).
2. Fenyö, E. et al. J. Virol. 62, 4414–4419 (1988).
3. Berger, E. A. AIDS 11 (suppl. A), S3–S16 (1997).
4. Moore, J. P., Trkola, A. & Dragic, T. Curr. Opin. Immunol. 9,
551–562 (1997).
5. Doms, R. W. & Peiper, S. C. (1997) Virology 235, 179–190.
6. Liu, R. et al. Cell 86, 367–377 (1996).
7. Samson, M. et al. Nature 382, 722–725 (1996).
8. Richman, D. D. & Bozzette, S. A. J. Infect. Dis. 169, 968–974
(1994).
9. Connor, R. I. & Ho, D. D. J. Virol. 68, 4400–4408 (1994).
10. He, J. et al. Nature 385, 645–649 (1997).
NATURE | VOL 391 | 15 JANUARY 1998